Tampere University of Technology

TUTCRIS Research Portal

Utilisation of generalised norms in the studies of Barkhausen noise measurement signals

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientific


Original languageEnglish
Title of host publicationThe British Institute of Non-Destructive Testing proceedings of 57th Annual Conference on Non-Destructive Testing
Subtitle of host publication10-12 September 2018, Nottingham, UK
PublisherThe British Institute of Non-Destructive Testing
ISBN (Electronic)978 0 903132 71 0
Publication statusPublished - 10 Sep 2018
Publication typeB3 Non-refereed article in conference proceedings
EventAnnual British Conference on Non-Destructive Testing -
Duration: 20 Sep 2018 → …


ConferenceAnnual British Conference on Non-Destructive Testing
Period20/09/18 → …


Industrial use of Barkhausen noise (BN) measurement signal uses predefined threshold values (usually the root-mean-square, RMS, value) in locating deviations within material. In research, more features are calculated and compared with material properties such as residual stress and surface hardness. Typical features include the RMS value and peak height, width and position. Generalised norms are computational features used in condition monitoring. In condition monitoring, also derivatives of signals are often studied instead of the raw signals. In this study, generalised norms of different order are computed from the BN measurement signal and its first and second derivatives. The
obtained features are compared with material properties. The traditional RMS is a special case of generalised norm and is used as the reference when analysing the results. The specimens studied are case-carburised hardened and ground. The results showed that the use of derivatives and generalised norms is beneficial when compared with the RMS value obtained from the raw BN signal. Thus the use of BN noise in applications could benefit if the derivatives and generalised norms are included in the analysis.


  • Barkhausen noise, generalised norm, Digital signal processing

Publication forum classification

Field of science, Statistics Finland